RE:On Just the Blockchain Vertical, DM appears well positionedThat article is interesting for more reasons than Visa's apparently inflated TPS figure. Putting that aside for the moment, the piece also indicates that some TPS rates trotted out by various blockchains are also inflated. Apparently they are able to "falsify" their TPS rates by "setting a lower block time and higher block size" or even "centralizing the authoritative nodes to avoid forks."
Now, I don't purport to understand all the technical intricacies of measuring blockchain transaction processing but this article certainly begs the question whether Graph Blockchain's 7000 transactions per second figure is legitimate or inflated.
How was Graph's "1000x faster" figure ascertained? The most recent presentation only says that "This
speed is powered by the highly performing engine and our unique approach" (Slide 12) and that the system "offers uncompromising performance thanks to its unique capability of data ingestion." (Slide 13.) However, nothing is really said about the technical specifics of this system nor even how the TPS rate was measured. What is the "unique" data ingestion capability? What is their "unique approach?" Seems very vague... no scratch that. It doesn't "seem" very vague. It IS absolutely very vague. Why? Are they using proper block time and block size settings to arrive at this 7000 TPS figure? Or are they, as the article states has been done by others, "taking the easy way out" and fudging/inflating the TPS numbers by tweaking the settings in the lab but could never reach those same rates in the real world? I don't know and I'm not making any allegations, but you should also know by now that I'm not afraid to ask the uncomfortable questions.
And before anyone starts yelling "trade secret" or "IP protection" remember, they already filed a patent application so there would be no harm or risk whatsoever in giving us more insight into how these TPS figures were derived.